Optimal Sleep Strategies with Respect to Traffic Patterns in Wireless Sensor Networks

Sleep strategies can be used in dense, redundant wireless sensor networks to increase network lifetime, but this comes at the expense of increased network management complexity. In this paper we consider the problem of queue stability in multi-hop networks with sleeping sensors. We derive stability constraints on a variety of sleep strategies as a function of network loading, channel noise, and data aggregation. We then compare and evaluate the performance of stable routing and scheduling policies over a set of relay nodes with and without queue length information. Through our analysis, we provide precise guidelines for the optimal design and operation of reliable sensor networks employing sleep strategies for a wide range of traffic patterns and applications.